Soldering Wires – Here's a different way
instructables.com
instructables.com
What you really want to do when you have dozens of jumper wires is to do an array-of-struct to struct-of-array transform on the steps. Instead of cutting, stripping, tinning, and soldering one wire at a time, cut all your wires, then strip them all, then tin them all, then solder them one by one.
Here's a pro tip: when you tin the wires, lay the iron flat on your workbench with tip out into air and tin by laying the end of the wire on top and then adding the solder.
Another pro tip: use a pair of needle nose pliers to bend the tinned leads by a little over 90 degrees to make a little hook. When you solder to the board you can apply a little tension and let the solder flow into the hole. Makes very string solder joints.
I call this "pin-making" after Adam Smith's famous/apocryphal account of the division of labour in a pin factory [0] and do it in pretty much every area of my life :)
[0] https://www.econlib.org/library/topics/highschool/divisionof...
You can even dip multiple wires into the flux and solder bath at the same time, as well as use the flux tub again before you solder the tinned wire to the contact.
Flux is probably more important than the solder for good joints. A syringe of paste flux makes life ridiculously easier.
What we did was use solder that wasn’t shitty. Big fan of felder 62/36/2 but it ain’t cheap.
We also cleaned stuff properly before soldering it. The reason people tend to flux things to death is that they didn’t clean the oxide layer off before soldering. Or used shitty solder.
This needs shouting from the rooftops.
So many people get sucked into the groupthink on this; yes it helps with less-than-perfect beginner's soldering skills, but it should not be normalized.
Flooding every joint with extra flux is unnecessary and a waste of money.
/Pet peeve (well, one of them!)
Most soldering that relies on effects from solder surface tension simply will not work without quite a bit more flux than you can get from the core of a solder wire. Drag soldering, for example, won't work on quad flat pack chips without quite a bit of flux.
And while my techs will absolutely back you up that my soldering skill sucks, to a person they ALL use solder flux for their soldering. Not even a single exception. And they're all trained to NASA aerospace standards.
Now, maybe you can be more skilled than they are and get away without using solder; however, back down here in reality, the rest of us punters will just add flux when soldering stuff, thanks.
It's also a leaded solder, which is much easier to get nice joints from than lead-free.
Flux is great, adding some before you solder is a good tip, and will likely help you make better joints if as you said, you haven't perfectly cleaned everything. The only downside is perhaps some more cleanup afterwards. There's no reason to disparage it's use.
0: https://www.felder.de/products/electronic-applications/manua...
Rework... good bloody luck.
Fluxless rework isn't difficult at all. I do it dozens of times a day.
Production reflow without extra added flux is normal. It's the goal for every assembly line anywhere and is almost always achieved.
Rework of existing joints after assembly without flux is just about impossible. The oxidation of the joints guarantees it. Many times the flux in the paste or wire is sufficient, but this is more true for through-hole than for surface-mount, mainly due to joint geometry and available volumes. If you are regularly reworking old fine-pitch surface mount joints previously assembled with no-clean flux, without flux of your own... then I'd like to know what materials you're using, because that is not possible in general. The "no-clean" fluxes are notorious for causing rework trouble, and they are difficult to clean off once reflow temperatures have been achieved (that's why we call them "can't-clean"). They are never active twice, so they must be removed and replaced once heated the first time. We have to use special solvent flux removers around here (based on DuPont's hellish Vertrel XF) to get the crap off just to be able to access joints again. (This is on assembly work done out-of-house or on COTS devices, made with who-knows-what.)
Nope, we don't use flux at all for rework. As the majority of our boards are solid metal, we just slap them on a heater and poke the components down or slide them into place. Flux isn't needed. Even with crappy RoHS tin solder. If you're getting significant-enough oxidation that interferes with rework with the product fresh out of the oven, or even a day or two later, you're doing something extremely wrong. Check and make sure your production paste isn't contaminated.
I really don't think either's wrong, and anyway I assume 'production engineering' is not using how-to-solder tutorials.
That is part of production engineering which is mostly about making sure that your product doesn't come back again because it's broken.
It's a different world from a hobbyist's one-off project. This advice was never intended for you, and appropriate advice can differ between audiences.
That said, I let someone borrow my bottle of liquid flux about a decade ago and have never bothered to replace it.
More flux that what comes with the solder isn’t necessary with good technique.
We're supposed to be using lasers to ... well, shoot at space ships, not cut plastic wire wrapping!!
EDIT - found the answer:
> Because the metallic conductor is essentially a mirror at the wavelength of the laser, the process is effectively “self terminating”, that is the laser vaporises all of the insulating material down to the conductor and then stops, so no process control is required to prevent damage to the conductor.
I'm imagining some researcher after years of investment and experimentation eventually accidentally discovering the solution: "you mean it just shuts _itself_ off?!?!".
If you want to laser through metals, you need a different type of laser. Just as getting through concrete needs a different tool (say, a hammer chisel).
A knife blade and a masonry chisel are roughly the same type of carbon steel tempered to a similar hardness.
They will both scratch and damage concrete.
But would you like to chop through my sidewalk to get to the water main with a santoku?
https://www.laselec.com/en/product/portable-laser-wire-strip...
Soldering stations aren't cheap (Hakko is my somewhat-budget-friendly favorite) but that (plus good solder) will make you realize you've been able to solder all along: it's just that you were using crappy tools.
I never really use the hot air part but am in love with the iron. I can dial in the temp and avoid breathing fumes directly. It is so much nicer to use than my old Weller with a temp dial.
Once I got a real iron, a real solder, and a real part holder my skills are infinitely better. I enjoy soldering now!
I almost exclusively use my Metcals now...the Aoyue has a tip for inserting heat set brass inserts to 3d prints nowadays, I don't use it to solder much.
Sidenote: My hotair station is almost 8 years old...still works fine. So for the money it was worth it for sure.
I dont do a ton but for me it always seemed like the weak point was the grip was 3+ inches from the tip. If I am soldering something that is just a couple mm from the hole next to it then this requires a decemt amount of dexterity - even if things are held nicely.
The nicer stations like Hakko's have very nice metallurgy in their tips as well.
As soon as you get a station too there's a little sponge, maybe some metal wool — these things you can use to clean the tip of the soldering iron. That in itself encourages good soldering hygiene and makes for better soldering.
Yeah my kit came up a sponge and station too - but unmounted was so lightweight it was worthless. Perhaps a safety threat but I (with my amateur perspective) still think a shorter length would be very beneficial if possible.
- solder fume extractor (see below)
- extra thin high quality tin
- a desoldering pump AND desoldering wire
- high quality flux
- a TS 100 soldering Iron with KU tip from ali express for 50 bucks
- optional: USB-C Power Adapter for your existing power supply (so you don't have to buy one)
- optional: a soldering mat (normaly blue colored)
Instructions:
- turn on solder fume extractor
- heat iron to 370° Celsius (but be careful depending on your workpiece)
- Tin both ends beforhand
- better slightly more flux to put on (but not for small works)
- if flux has been applied, cleaning up with e.g. toothbrush & isopropyl alcohol after you're done maybe a good idea
High quality flux is also important because of your health. Chinese sellers often don't care about ingredients...
High quality tin is much easier to work with and costs 5 bucks more than low quality tin. It is worth.
I got to here and thought this was some elaborate joke, and expected the github page to continue on with the joke. An operating system for a soldering iron?? But no, it seems 100% serious. Somehow soldering iron tech passed me by.
It's the flashlight guys who have really run amok:
https://www.reddit.com/r/flashlight/comments/xopqft/prob_a_g...
Just joking... mostly.
After reading this thread, it sounds like I need to get myself a better soldering iron and install IronOS
The "OS" part is a bit of a joke. It's just an embedded application, AIUI they didn't write their own RTOS.
Plus, it's running Risc-V, and they even sell an breakout to use the "logic board" as a RISC-V dev-board!
[1] https://pine64.com/product/pinecil-smart-mini-portable-solde...
The TS100 was about 40$, when I bought it...
What advantage, if any, would the ts-100 have?
Anyway, USB-C is a big deal for me so I'd take the Pinecil if prices were the same.
https://pine64.com/product/pinecil-smart-mini-portable-solde...
Full disclaimer I know some people had trouble with the USB-C power input so YMMV, but considering the price it's probably the most cost-effective tool I ever bought.
Oh I should mention I have both the TS100 and Pinecil, and the Pinecil does everything the TS100 does.
Don't take my word for it, it's all explained in the NIOSH and OSHA handbooks.
Safety is not priceless. Regulators carefully calculate how much impact various safety equipment and practices have relative to their cost. And you too should consider how much you need it. Solder for hours a day, get the best fume extractor you can find, do a hobby project every other month, you can go cheap.
Anyways, I wonder if a simple small desk fan fitted with a surgical mask you probably have too much of would be a good compromise of simple, cheap, and better than nothing for hobbyists.
If you're fixated on getting yourself a fume extractor... well, to each their own.
I took a look at it, ultimately got a TS80P. It's smaller, and fed via USB-C. Seems to work perfectly fine.
> solder fume extractor
How big of a deal is solder fumes of unleaded solder?
Higher temp
lighter
faster heat up
easier temp adjustment
> How big of a deal is solder fumes of unleaded solder?
Depends on solder, but spending $40 for not risking your health is affordable in my opinion...
Soldering temperatures will not vaporize lead or the metals in unleaded solder.
There is also some conjecture that the fluxes used in unleaded solder produce more aggressive fumes.
Worse than leaded solder, because the temperatures are higher.
The lead in leaded solder doesn't go up in fumes. Contamination is from touch. You should wash the board after you're done (with e.g. IPA) and your hands.
You also should try and not solder where you eat, and the other way around.
Because it's a good iron. So is the TS80p.
Pine64 has the Pinecil, a TS100-compatible improved replacement that is RISC-V based.
Buy a decent iron (metcal - get a second hand one off eBay) and some decent solder (felder 62/36/2) and you don’t need it.
Also don’t use a toothbrush with isopropyl. They tend to dissolve and leave crap all over the board.
Toothbrushes bristles may melt. Not all of them are nylon and if they aren’t they will go sticky and this will wipe off on the surfaces.
Hobbyists and technicians get obsessed with cleaning off flux. Hobbyists have no business cleaning flux off boards. It's very easy to wind up freeing corrosive salts from the rosin matrix. Then they smear all over the board to corrode and conduct. My rule of thumb is if you can't measure the cleanliness of your board, don't clean it.
> felder 62/36/2
Other readers please note that felder 62/36/2 has a flux core. Which isn't clear in this comment.
If you are in the basement without window, you probably need a better one with HEPA + active carbon.
I would say for hobbyists you should spend around 20 - 50 bucks for a used one.
Maybe also care about the loudness... these things are really bad ;)
I ordered mine as used device here in Germany. Aliexpress would have taken weeks and Amazon was just to expensive (cheapest about 80 bucks).
This sounds like way too much. Yes, with an iron you don't have the luxury of going just above melting point. It's a balance between heating up workpieces quickly (so that heat does not conduct too far) for what you need higher temperatures, not burning the flux and solder solidifying as quick as possible, because any movement in solidifying solder WILL reduce quality of the solder. Lower temperatures are almost always better than high temperatures.
You are better off getting powerful iron with good temperature control rather and keeping temperatures in working range rather than heating iron to very high temperatures and having it cool down during soldering if the iron cannot keep up with cooling. For soldering wires to a breadboard it may not be much of a problem (sans insulation burning off), but I would not risk damaging silicon with temperatures that high.
With 60/40 solder temperatures around 300 [290-310] are usually more than enough, depending on workpiece size.
Also, that iron looks like a Hakko "936" cheap clone: they are barely above the garbage level; get a better one. Used Wellers aren't that expensive, and no, you won't need a microcontroller for digital temperature control and all those gimmicks they put in modern solder irons. I also have a PineCil from the same folks that made the PinePhone and was amazed by its quality. Yes, it has a mcu for digital temperature control, but I'd lie if I wrote that I used any of its functions beyond setting the temperature.
I've liked the Pinecil except sometimes mine crashes in the middle of me soldering. I need to update the OS but I'm lazy. It does fail cold (Says it's hot but it's not heating) which is probably the better of the two ways it could go.
A lot of technicians seem to like to have temperature knobs so they can dial the temperature around instead of swapping out for an appropriately sized tip: this works of course except when it doesn't.
STANDARD MAINTENANCE PRACTICES MINIATURE/MICROMINIATURE (2M) ELECTRONIC ASSEMBLY REPAIR
https://www.robins.af.mil/Portals/59/documents/technicalorde...
I was a solder tech in the Navy. That was our Bible.
I wonder which ones came first given how much aerospace is miltary work.
The final straw was when I saw someone using a wire brush on a Metcal tip after reading one of these mil spec guides.
I was production engineer defence sector for a few years.
We worked in R&D and all (Multibus II) boards were hand-soldered and had to be visually inspected by a supervisor before testing.
One chap presented his board for checking and, after a visual and touch inspection, our supervisor said that some of the joints were a bit 'pointy' and would need fixing.
So off went the student to the fab area of the workshop, the board was put in a vice and I just stopped him as he approached the bench with a sandpaper-wrapped block of wood.
This looks like they are laying the wire on the solder, then heating the solder - not the wire. Won't this result in a cold solder on the wire?
The bigger problem is that all the rosin flux in the solder will have evaporated, likely making the joint very weak. Most DIY projects lack any form of strain relief so it's a double whammy for any wire that will shift around.
But the OP specifically adds short pieces of solder to the pads before soldering wires. When they finally lay the wire and start heating that solder piece might have just enough flux in it. Alternatively one could just smear some flux on the pad before laying wire.
Still I am not convinced this particular way is of any practical advantage.
But yeah, my solution is to just dip the wire in a jar of flux as I’m stripping and twisting them
Never put solder on the iron tip...although...I do it sometimes to increase heat transfer to larger joints (emulates a larger tip) but you can expect that the flux has burned off and may require liquid flux and reflow.
Sidenote: if you coat your tips with solder before shutting off your iron it will prevent oxidation and increase tip life.
You should avoid soldering when possible because connectors are detachable for module maintenance and solder is prone to cracking in high vibration environments. Typically you will see a connector soldered and otherwise secured mechanically to a PCB where vibrations are unavoidable.
Crimping is better whenever feasible
Most people following Instructables for tips on soldering wires to vector boards aren't likely to need to maintain their prototypes, and roughly 0% of those prototypes will be in a high-vibration environment.
Anyone who has debugged one of them knows to do things properly or suffer the wrath of intermittent problems.
For wires like the ones in the video you can still use something like a small phoenix connector and just screw your wire connections in, just like you would wire up a choc box, no crimp tools required.
The advice in the post will likely create a cold joint and intermittent/noisy connections.
I started out soldering but now it's mostly limited to installing connectors into PCBs followed by hot glue around the housing for mechanical rigidity
If you're an individual making runs of 1 and you have to crimp a different connector to a different wire every time, soldering is easier and more likely to result in a reliable connection.
For prototypes or one-offs, soldering as you say is more likely to be reliable.
It’s actually much harder to make a good solder joint with lead free solder, so be kind to yourself about it. No harm in re-doing a joint.
Anyhow, I never fried a part with the iron. Heated the wire just enough to flow the solder, and applied heat only barely long enough.
To avoid needing a third hand for some joints, I'd pre-fill the hole with solder. Then heat it up again till it melted, and push the wire in.
Once I switched to Kester solder, I had a much better more consistent experience. My hypothesis is that cheaper solder doesn't have the flux as evenly distributed, so you get spans of solder with too little flux and then everything gets worse, then you hit a good patch and it gets better.
P.S. I worked my way through college being an electronics assembly technician, so I spent a lot of time soldering. I enjoyed making the boards look perfect.
It turns out, this skill is transferable to soldering copper water pipes. Never had a leak! (Using a torch not an iron.)
I didn't go lead-free because I've heard leaded solder is easier to use (lower melting point?) and since I'm just a part-time hobbyist, the health and environment considerations are minimal. It's not like I'm soldering eight hours a day every day.
IMHO...despite the risks of lead...leaded solder is way easier to use and especially inspect. Lead free often looks "cold" just by the color of it.
Use a larger tip for higher heat transfer. Don't increase temp...this is how people burn boards and lift pads.
You can solder on 0805 parts and switch to soldering bolts onto nuts without changing the tip or any settings.
If you have the power without the thermostat, you will reach unsafe temperatures and burn what you're not supposed to burn.
PID controllers make a huge difference.
(Through holes are fine. Mostly.)
For quick one offs, simply place a normal pad, place a copper fill polygon connected to the same net around the pad, and turn off thermal reliefs / set pad connection to solid for the polygon. Obviously if you need it a lot, draw a new footprint.
Just stay away from any Molex crap. Tends to charcoal itself.
Solid wire through holes is OK if the wire is never going to flex. Otherwise the stranded wire lapped to surface mount pads, or soldered into through holes, with a big blob of hot glue or epoxy for maybe-sorta "strain relief" is a great pet peeve of mine - it's ugly and it breaks.
Never had a problem with Molex, myself.
You need to nail the pad down with a few vias (under soldermask is fine) or use a through hole. Please just use a through hole if you can. They're a lot more reliable than just about anything else. Even SMT connectors rip off boards without much trouble. (This is my job. I have seen some things.)
But it’s surprising how poor the reliability can be. Aside from your points, solder can wick up a cable, and then the cable often breaks just at the point where the solder ends since it has no strain relief and may be right in the middle of a bend. This is especially true for adding solder to a crimped connector “for good measure”. It does more harm than good.
That being said it’s convenient many times especially if you don’t have connectors on hand or need to place wires in a specific orientation. Hot glue or better yet, RTV silicone can be used to make strain relief to increase the reliability. Or, when using perf board like in the post, a good technique is to drill a hole for a snug fit of the cable including insulation, then pass the cable through the hole before soldering to an adjacent hole.
It’s worth expanding on my comment now I have a few minutes. You can terminate solid core like this if it has some strain relief but not stranded!.
My favoured construction method is to take a Hammond box lid and screw a piece of blank FR4 single sided copper to the inside of it leaving enough space around the edges for any (proper) connectors you need. You can then drill out anything you need, usually BNC, SMA, DC jacks, FT capacitors and D connectors for me. You build the prototype on the FR4 board dead bug and use solid core or RG174 to wire up the connectors and lap solder them to the boards or components as required. This is done carefully with thought for weight distribution and to keep the total mass of the wires low. When it works the Hammond box is closed. If it needs to withstand vibration or moisture you build it into the bottom of the Hammond box and when done fill it with potting compound.
You MAY have a chance if the wire is pretinned or you use flux. Do not do this with bare copper even if it looks like it works, you will probably get at least a few cold joints.
The better version is to coat the wire and pad with solder, put the wire on the pad, flux. then press it into the pad with the iron.
Better yet, never ever solder wire directly to a board in hobby or low volume work unless small size is totally critical. Especially stranded. It makes a metal fatigue point if things move at all, and requires hot glue to stand even a few bend cycles.
And even when you do have good connectors, it's probably still in your best interest to minimize the number of wires hanging around, learn about I2C as quick as you can.
It can't be quickly disconnected for repairs and replacement and troubleshooting.
It's a lot of work.
Even if done perfectly it's just not very nice.
Instead, solder connectors to the board directly. For short ranges use 2.54mm jumpers. Never use male jumpers if you can avoid it, they are delicate and annoying.
For longer distances, there are lots of things that don't suck, USB-C, 2.1mm barrel, Ethernet, 3 pin XLR, and whatever your country has for mains electricity.
Stay away from normal 4mn banana plugs unless you are in a field where they are already everywhere. The cheap ones shear off. and have bad connections and they are not particularly common. Even DMM test leads don't use them, they use the shielded versions.
Common off the shelf stuff that you can buy extensions for. Don't solder long wires to stuff, it will just make a tangled mess when you put it away and make it hard to troubleshoot.
When you do have a reason to do stuff like that, Wago connectors and pigtail adapters are your friend especially for tests and troubleshooting Because then when you realize you actually don't need 5 XT60 extensions, you can swap the ends for something else quickly.
And most importantly Do. Not. Invent. Some. Crazy scheme of powering all your projects from a central point over long 12v wires. Every new electronics DIYer seems to do it, and cables and wires are often the enemy. Yeah it looks cool and sci-fi and seems like a good idea to have wires everywhere, but single purpose and custom made cables are a nuisance to deal with, they're heavy and expensive and trip hazards.
My trick is to tin the joint and the wire, apply no-clean flux to either the joint or the wire (situation dependent), then bring the two together while reflowing the joint. Quite handy for soldering tight multi-pin connectors.
The real tip is buy flux, it’s just as important as solder.
The other way you can easily solder to pads is to use solder paste. Many people don't seem to realize that paste is good for more than just reflow. It's not as useful for situations like this, where you need a relatively large solder volume (so I'd also use your method), but it works quite well for smaller scale work.
One of the images shows damaged insulation and too much solder. https://content.instructables.com/ORIG/FES/OFV3/KZ7C3ZX0/FES...
For actual chips, SPI programmer clips have been great. For test pads, I just ordered one of these [1], which will hopefully work just a well.
Interesting the link you posted - it seems 2x5p 1.27mm version with one pin cut would work as the pogo version of Hack Connect.
For me, I struggle to get solder even in the best of circumstances. It's just a skill that I cannot master. I really, really like this technique as it enables me to make passable hobby solders without risking messing up my project.
The only "good" I see in this is that they author has left a rat-tail of unmelted solder so at least it will have some flux left to wet the wire.
* which show up as intermittent opens at DC with some fun dependence on position / temperature / humidity. At audio frequencies, they're disastrous -- I can only imagine what they'd do at HF.
I'm not saying that I celebrated the moment when one of my wires or jacks broke, but it felt great fixing it with a little patience and some soldering.
I am close now to many musicians that never fixed any of their wires/cables.
I kinda like to belong to those older times.
I run my iron really hot (750F/400c) and just get in and out. I use a curved conical tip and have never have had issues with this technique.
You do have to wet your tip more frequently when you are using high temp with flux core. However, I end up with shiny, well made joints 99% of the time. If I burn off the flux, I typically will wick it up and re-apply, or just swipe the solder off the joint and add more solder (depending on what I am working on).
I know there are aficionados who can point out possible disadvantages of this approach, but honestly, I can see this technique being really handy when you just want to prototype something out and you want to get things done efficiently. Not every solder joint needs to be perfect, just as not every PCB needs to be factory quality and is going to a customer. As someone who only occasionally needs to solder a custom circuit, this looks like a time saver. If I was frequently making PCBs by hand, maybe I'd think otherwise.
In my first two attempts, I managed to swap pickups successfully, where "success" is measured as "it sounds the way I expect, all knobs and switches work, there is no unexpected buzzing." I worry about how much I don't know about it, so I'm curious: is it possible that I need to expand my concept of success? Should I be concerned that a joint that's fine today won't be OK tomorrow? Or is it enough to say that if it sounds right and doesn't look like trash, it is good enough?
It depends on the application. If you're not performing, the cost of a joint that fails later is just a bit of annoyance. If you're on stage and your equipment fails, that's a lot worse.
I wouldn't think guitar solder joints are stressed out too much though, so if it lasts a week, it'll probably last longer. There's a little bit of vibration and motion, but not a lot of power or heat. And most people try to take it easy on their guitars --- it's not like car wiring or anything.
Yes. Things can go wrong with solder connections over time, such as:
1. So called "cold" solder joint eventually becomes non-conducting or semiconducting (interesting properties of tin) over time, resulting in weird behaviors such as AM radio pickup.
2. Effects of vibration. The guitar is probably getting handled a lot.
3. Other stuff coming loose, such as the nuts holding pots and jacks in place, causing wiggling of the solder joints.
4. Not directly related to soldering, but poor stripping can "nick" the conductors of the wires, leading to premature breakage.
I'm an electric bassist, and I'm responsible for my gear being 100% reliable on the bandstand. I don't bring a spare bass, except that my electric is sometimes a spare for my upright, especially for outdoor gigs. I do keep some spare cables and other odds and ends in a bag that stays in my car. I'm also quite experienced with electronic tech work.
There are some things that can give you a better than 0% chance of your stuff staying in business over time:
1. A good wire stripper that has fixed blade settings and can't nick the wire, such as a "T Stripper" or an "Ideal Stripmaster."
2. A commercial quality iron with a tip in good condition.
3. Lead based solder is still easier for beginners, but a lead-free alloy called Kester K100 is a lot better than the early lead-free alloys.
4. Make sure both sides of the joint are "tinned" before joining them together and soldering. Then you can complete the joint with just a touch of the iron and a bit of solder. For weird / old components, tinning is a good chance to make sure that the solder will actually wet the component, as opposed to balling up.
5. Strain relief. Wire runs of more than a couple inches should be tied with with zip ties or whatever.
6. Be on constant lookout for deterioration of your instrument, something that's just got to be part of being a musician. If a jack or pot comes loose, tighten the nut. You can sneak under the knobs with a bicycle cone wrench if the knob is hard to remove. You'll need the cone wrenches for your bike too.
and as many have already said, pretty good way to get a cold joint or pockets
Like the others here, I'm not convinced this is better in any way than the "normal" method, or just using solder paste.